Beam-based Feedback Systems

Transcription

1 Beam-based Feedback Systems Philip Burrows Queen Mary, University of London

2 ILC Beam-based Feedback/Feedforward Systems Intra-train (bunch-bunch) feedback at IP: 3 MHz Pulse-pulse feedback at IP: 5 Hz Orbit feedbacks upstream in BDS + in linac: < 1Hz Corrections to beam position and/or angle IP FBs: most critical in vertical dimension in order to make nanometre-sized beams collide Feed-forwards: DR linac, linac IP

3 Critical Feedback/Feedforward Components BPMs BPM signal processor Feedback processor Amplifier Kicker Intra-train FB at IP:

4 IP Intra-train Beam Feedback For BPM (say) 4m downstream of IP: for beam offsets < 50 sigma_y beam position signal < c. 1mm BPM resolution of order 10 microns is adequate stripline or cavity BPMs ok Kicker needs to provide up to 100 nanoradian deflections cavity BPM needs <<300ns time resolution amplifier power < 100W

5 Intra-train Feedback Performance (White) 3 x 1034 Luminosity / cm -2 s -1 1 seed: post-bba + GM + wakes 2 1 y position FB: restore collisions within 100 bunches y position scan: optimise signal in pair monitor y angle scan Bunch # 10 5 OPTIMAL LUMINOSITY µ = ± Fractional Lumi Change After IP FB 15 5 ILC IR1 (20 mrad) µ = ± Fractional Lumi Change After ANG FB

6 FONT Prototype Intra-train Analogue Feedback Systems NLCTA: 65 MeV beam, 170ns train, 87ps bunch spacing FONT1 (2001-2): First demonstration of closed-loop FB: latency 67ns 10/1 beam position correction FONT2 (2003-4): Improved demonstration of FB: latency 54ns real time charge normalisation with logarithmic amplifiers beam flattener to straighten train profile solid-state amplifier ATF: 1.3 GeV beam, 56ns train, 2.8ns bunch spacing FONT3 (2004-5): Ultra-fast demonstration of FB: latency 23 ns 3 stripline BPMs high-power solid-state amplifier

7 FONT: kicker driver amplifiers FONT1 3-stage tube amplifier FONT3 PCB amplifer + FB Same drive power as needed for ILC

8 FONT3 outline Adjustable-gap kicker BPM ML11X BPM ML12X BPM ML13X Superfast amplifier Superfast BPM processor Aim: Feedback TOTAL latency < 20 ns

9 FONT3: BPM processor + amplifier/feedback installation in ATF beamline BPM processor board FEATHER kicker Amplifier/FB board BPM

10 FONT3: latency budget Time of flight kicker BPM: 4ns Signal return time BPM kicker: 6ns Irreducible latency: 10ns BPM processor: 5ns Amplifier + FB: 5ns Electronics latency: 10ns Total latency budget: 20ns Will allow 56/20 = 2.8 periods during bunchtrain

11 FONT3: BPM processor tests (Molloy) (single-bunch, December 2004 beam tests)

12 FONT3: BPM processor latency measurement (single bunch, March beam tests) latency Distribution of latency: processor output raw signal Latency 4.3 ns

13 FONT3: BPM scale calibration using correctors (20-bunch data, March beam tests) BPM12 BPM11 BPM13

14 FONT3: BPM position resolution (20-bunch data, March beam tests) Distributions of residuals (240 beam pulses): BPM11 BPM12 BPM13 Resolution: 3-5 um

15 FONT3: Results (June ) 40 pulses per position setting

16 FONT3: Average over 40 Pulses 56ns bunchtrain 200um FB on 23ns FB + delay loop on

17 FONT3: Summary Demonstrated feedback with delay loop Ultra-fast system: total latency 23 ns Varied main gain, delay loop length, delay loop gain - system behaves as expected Beam quality + limited time (6 shifts) did not allow detailed optimisation of system parameters

18 FONT1,2,3: Summary 67 ns 54 ns 23 ns Even fast enough for CLIC intra-train FB!

19 FONT4: Digital FB Processor Module (Dabiri Khah) JTAG connector JTAG circuit RAM Flash/ EEPROM DAC AOUT2 DAC AOUT1 Serial connector UART circuit FPGA ADC Differential To single AIN4 USB connector IN Power Jack & switch USB circuit 5v 3.3v 2.5v? v Clock circuit Clk IN ADC ADC ADC Differential To single Differential To single Differential To single AIN3 AIN2 AIN1 Latency goal 100ns

20 FONT4: Prototype Digital Feedback System ATF/(ATF2): 1.3 GeV beam, 3 bunches with spacing c. 150ns FONT4 (2005-6): modified FONT3 BPM front-end signal processor digital FB processor modified FONT2 solid-state amplifier: 300ns long o/p pulse FEATHER adjustable-gap kicker Aiming for first demonstration of FB w. ILC-like bunches: latency 100ns (electronics) stabilisation of 3 rd bunch at um level Possible first component tests at ATF December 2005/March 2006

21 5-Hz Integrated Feedback Simulations (Linda Hendrickson) Linac feedback distribution: 5 distributed loops per beam, each with 4 horizontal and 4 vertical dipole correctors, and 8 BPMs (X&Y). LJH simulations. BDS feedback distribution: 1 BDS loop per beam, 9 BPMs and 9 dipole correctors, both horizontal and vertical. Based on NLC simulations by Seryi. Linac and BDS feedbacks Cascaded system of 6 loops per beam: loops don t overcompensate beam perturbations, but can be independently disabled for operational convenience. SLC-style single cascade (each loop communicates beam information to single adjacent downstream loop). Linac and BDS loops have exponential response of 36 5-Hz pulses. IP deflection (X&Y), not cascaded, exponential 6 pulses (like SLC). Matlab/liar/dimad/guinea-pig platform. Upgraded liar/dimad for energy and current jitter, and dispersion measurements. KEK-model ground motion (noisy site).

22 Feedback Simulations (Linda Hendrickson) BPM readings after 30 minutes ground motion Emittance growth in linac 5 distributed linac loops ~100% after 30 min KEK ground motion + jitter for 10 seeds, 6% with feedback (3% with feedback without jitter). 1 BDS loop IP loop

23 Feedback Simulations (Linda Hendrickson) Banana-bunch shape is seen at end of LINAC after 30 minutes of K ground motion. Fixed with feedback.

24 Single-beam studies of beamsize growth, with 5-hz feedback in LINAC and BDS. Perfect initially, add 30 minutes KEK ground motion, let feedback converge -> 5% beamsize growth (380% without feedback). Increase energy spread for undulator (.15% end of linac; this effect needs more study!) -> 14%. Beamsize growth effects, with feedback (Hendrickson) 30 min ground. + Undulator + Component jitter + 5 Hz ground. + Kicker, current, energy jitter, BPM resol. Add component jitter (25 nm BDS, 50 nm linac) -> 15%. Add 5-Hz KEK ground motion -> 18%. Add kicker jitter (.1 sigma), current jitter (5%), energy (.5% uncorrelated amplitude on each klys, 2 degrees uncorrelated phase on each klys, 0.5 degrees correlated phase on all klystrons, BPM resolution.1 um. ->

25 2-beam Integrated Feedback Simulations (Linda Hendrickson) 2 beams, 5-Hz linac, BDS and IP deflection feedback. Perfect initially, feedback turned on after 30 minutes of KEK ground motion. 5 Hz ground motion, added component jitter, kicker, energy, current jitter. No angle feedback, no intratrain feedback. For the first ~20 seconds, IP feedback cannot keep up with large BDS steering changes. After 20 seconds, beams kept in collision but luminosity is poor (~20% in preliminary simulations, ~79% with perfect intratrain IP feedback).

26 International Fast FB Collaboration FONT: Queen Mary: Philip Burrows, Glen White, Glenn Christian, Hamid Dabiri Khah, Tony Hartin, Stephen Molloy, Christine Clarke, Christina Swinson Daresbury Lab: Alexander Kalinin, Roy Barlow, Mike Dufau Oxford: Colin Perry, Gerald Myatt SLAC: Joe Frisch, Tom Markiewicz, Marc Ross, Chris Adolphsen, Keith Jobe, Doug McCormick, Janice Nelson, Tonee Smith, Steve Smith, Andrei Seryi, Mark Woodley, Linda Hendrickson. FEATHER: KEK: Toshiaki Tauchi, Hitoshi Hayano Tokyo Met. University: Takayuki Sumiyoshi, Hiroaki Fujimoto Simulations: Nick Walker (DESY), Daniel Schulte (CERN) Eurotev (BDS, ILPS), ELAN (Instr, Bdyn), ILC WG4 (BDS)

27 Spare slides follow

28 FONT3: Averaged results (HIGH gain, nominal delay settings)

29 FONT3: Averaged results (LOW gain, nominal delay settings)

30 FONT3: Average results (variation of delay-loop settings) Delay loop length Delay loop gain